The present invention relates generally to a new apparatus and method for introducing thermal paste into semiconductor packages. More particularly, the invention encompasses an apparatus and a method that uses at least one preform of thermal paste for the cooling of at least one chip in a sealed semiconductor package. The thermal paste preform is subcooled, and is transferred onto a module component from a separable transfer sheet, or is placed onto the module component using an attached and/or imbedded mesh. The preform of thermal paste may be of simple or complex shape, and enables cooling of one or more chips in a module.
Semiconductor devices are becoming smaller and more dense with the evolution of new technology. However, increases in circuit density produce a corresponding emphasis on overall chip packaging strategies in order to remain competitive. Chip and substrate manufacturers are therefore constantly being challenged to improve the quality of their products by identifying and eliminating problems, reducing package size and weight, decreasing package costs, providing improved thermal efficiencies and better and more advanced chips. Whereas significant improvements are being made to eliminate systematic problems by reducing process variability, process improvements alone are not sufficient to eliminate all the problems which effect both performance and reliability.
One way to allow high degrees of integration is to provide a highly efficient internal cooling design. A preferred way of cooling high performance SCMs (single chip modules) and MCMs (multi-chip modules) is using thermal paste. Thermal paste is often used as a high thermal conductivity interface material to fill the gaps between the back-side of chips, such as, flip chips, and the inside surfaces of the caps or heat spreaders. Producing modules that use thermal paste has multiple challenges. Module components, like the back surface of the chip and the inside of the cap must be chemically compatible with the thermal paste, so that the paste can adhere to them. The package must be designed such that the thermal paste filled chip-to-cap gap has sufficient thickness that it will form a reliable structure.
During assembly, sufficient paste must be dispensed to completely cover the chip. Care must also be taken to use a dispense pattern that will cause the entire chip to be covered. Thermal paste is often very viscous and difficult to handle or dispense. The paste dispense may be accomplished by either use of expensive automated equipment, or by manually screening the thermal paste through a template.
U.S. Pat. No. 3,993,123 (Chu et al.) discloses the thermal conduction module (TCM). This was a major advancement in the cooling of MCMs. It utilized a complex cooling hat. Over each of the flip chips on the substrate was a hole in the hat. From each hole extended a spring loaded piston that contacted the back of the chip. The module was hermetic, and filled with helium. The primary cooling path was from the circuit side of the chip, through the thickness of the chip, to the face of the piston, up the piston, to the inside of the hat, to the back of the hat, across an interface to an attached cold plate, and to the water circulating through the cold plate. The high helium content of the gas in the module greatly reduced the thermal resistance of the chip-to-piston interface and the piston-to-hat interface.
U.S. Pat. No. 4,193,445 (Chu et al.) discloses another enhancement on the TCM. Here solder was included with each of the pistons so that it could be reflowed after assembly, to fill the chip-to-piston and piston-to-hat gaps, for improved thermal performance.
U.S. Pat. No. 4,226,281 (Chu) disclosed an enhancement to the TCM. Rather than just one piston per chip, each chip now had a matrix of pistons that contacted the back of the chip for cooling. To maintain almost full coverage of the back of the chip, headers are used on the faces of each of the pistons.
U.S. Pat. No. 5,005,638 (Goth et al.) disclose an improved piston geometry. Barrel shaped pistons were used to allow tighter piston to hat gaps, while maintaining the ability to accommodate chip tilt. Material changes also improved thermal performance over the traditional TCM. Pistons were now made of copper rather than aluminum, and the module was now filled with oil rather than helium.
U.S. Pat. No. 5,023,695 (Umezawa et al.) discloses a flat plate cooling (FPC). In this structure, a flat cooling plate is just above the array of chips. Thermal paste is used to fill the gaps between the chips and the flat plate.
U.S. Pat. No. 5,098,609 (Iruvanti et al.) the disclosure of which is incorporated herein by reference, discloses stable high solids, high thermal conductivity pastes. The pastes include a thermally conducting solid filler, a non-aqueous liquid carrier and a stabilizing dispersant. The resulting pastes are highly concentrated, of low viscosity, electrically resistive, highly thermally conducting and stable.
U.S. Pat. No. 5,291,371 (Gruber et al.) disclose a thin high performance cooling structure. The purpose of the structure is to establish and maintain a thin solder interface between the chips and the cooling hat. In order for this to be functional and reliable, the solder and the hat have to be very flat, and a thin lubricant is between them. The solder can be positioned onto the chip as a preform (and then reflowed or plastically deformed).
U.S. Pat. No. 5,325,265 (Turlik et al.) discloses a higher performance version of flat plate cooling. In this invention thermally conductive cushions conduct heat from the backs of the chips to the inside surface of the hat. The hat has shallow cavities above each of the chips. The cushions are low melting point solder, preferably indium. The indium may be placed between the chips and the hat as preforms.
U.S. Pat. No. 5,591,789 (Iruvanti et al.) the disclosure of which is incorporated herein by reference, discloses a polyester dispersant for use in high thermal conductivity pastes.
U.S. Pat. Nos. 5,604,978, 5,623,394 and 5,724,729, (Sherif et al.) the disclosure of which is incorporated herein by reference, disclose a method and apparatus for the customized cooling of chips on an MCM with a range of cooling requirements. It uses flat plate cooling, and uses pastes of different thermal conductivities on chips to customize the cooling of the chips. The paste is either between the chips and a flat cooling hat, or between the chip and a blind hole in the hat. Surplus paste may also fill some or all of the rest of the inside of the module.
U.S. Pat. No. 5,718,361 (Braun, et al.) the disclosure of which is incorporated herein by reference, discloses an apparatus and method for forming mold for metallic material. Braun also discloses the use of a mold to form heat sinks with fins.
U.S. Pat. No. 5,718,367 (Covell, et al.) the disclosure of which is incorporated herein by reference, discloses a mold transfer apparatus and method. Covell also discloses the use of a mold transfer to form a high density heat sink.
U.S. Pat. Nos. 5,757,620 and 5,819,401, (Edwards, et al.) the disclosure of which is incorporated herein by reference, discloses a method and apparatus for cooling of chips using blind holes with customized depth. It uses flat plate cooling, and varies the depth of the thermal paste filled gap to customize the cooling to each of the chips on a module, such as, a MCM (Multi-Chip Module).
U.S. patent application Ser. No. 08/758,789, filed on Dec. 3, 1996, now U.S. Pat. No. 5,85,087, and U.S. Pat. No. 5,770.478, (Iruvanti, et al.) the disclosure of which is incorporated herein by reference, discloses a hermetically sealed module where the internal surface of the module has a roughened surface by grit blasting or machined to have parallel and/or crossing grooves. The paste penetrates the roughened surface and inhibits the flow of the paste out of the gap.
U.S. patent application Ser. No. 09/140,583, filed on Aug. 26, 1998, now U.S. Pat. Nos. 6,111,314 and 6,255,139, (Edwards, et al.) the disclosure of which is incorporated herein by reference, discloses a structure and a method that uses surface chemistry modification of the inside of the thermal cooling cap where it contacts the thermal paste. This is done by modifying the internal surface of the cap by embedding particles that preferably have the same chemical composition as one or more of the solids in the thermal paste.
As stated earlier a variety of internal cooling means have been disclosed, using complicated hardware designs, helium filled modules, solder interfaces, thermal paste filled interfaces, etc. Thermal paste offers the combination of low cost hardware, high thermal performance, and module reworkability. What is needed are improved methods of introducing thermal paste into semiconductor modules.
One way to improve the application of thermal paste in semiconductor modules would be to use thermal paste preforms. The preforms can be formed with conventional thermal paste, and then subcooled to temporarily increase stiffness and decrease tackiness. The thermal paste may be on a transfer sheet, or attached to a mesh and would be transferred onto or associated with the chip that needs to be cooled.
The invention is a novel method and an apparatus for making and using at least one preform of thermal paste for the cooling of at least one chip in a capped module.
Therefore, one purpose of this invention is to provide an apparatus and a method that will allow the making and using of at least one preform of thermal paste for the cooling of at least one chip in a capped module.
Another purpose of this invention is to provide thermal paste preforms that are of simple or complex shapes.
Yet another purpose of this invention is to provide preforms of thermal paste that are used to cool one or more chips in a module.
Still yet another purpose of this invention is to provide thermal paste preforms that are transferred onto modules from separable transfer sheets, or are placed onto modules using an attached and/or imbedded mesh.
Still another purpose of this invention is to provide thermal paste preforms to enhance the thermal path between a chip and a heat sink.
One advantage of the current invention is that the thermal paste may be incorporated into the module without the requirement for expensive automated tooling.
Another advantage of the current invention is that complex templates are not required.
Yet, another advantage of this invention is the improved control of the quantity of paste applied to each chip.
Still, another advantage of this invention is the ability to use complex patterns.
Still, yet another advantage of this invention is the improved reliability, when the mesh is used and is left inside the module.
Another advantage of this invention is eliminating the need for manual screening of paste.
Yet another advantage of this invention is the reduced cycle time for assembly of a module.
Therefore, in one aspect this invention comprises a heat transfer structure comprising at least one preformed slug of at least one thermal paste sandwiched between at least one chip, acting as at least one heat source, and at least one thermal dissipation device.
In another aspect this invention comprises a heat transfer apparatus comprising at least one heat generating device with at least one thermal paste preform secured to at least one surface of said at least one heat generating device.
In yet another aspect this invention comprises a process for providing heat transfer to at least one chip, comprising the step of placing at least one preformed slug of at least one thermal paste on said chip to provide said heat transfer.
In still another aspect this invention comprises a process of forming at least one preformed slug of at least one thermal paste, comprising the steps of:
(a) filling at least one mold with at least one thermal paste,
(b) subcooling said mold and said thermal paste to a temperature that temporarily stiffens and reduces the tackiness of said thermal paste to form said at least one preformed slug, and
(c) removing said formed subcooled preform slug from said mold.